Ta. Mousseau et Da. Roff, GENETIC AND ENVIRONMENTAL CONTRIBUTIONS TO GEOGRAPHIC-VARIATION IN THE OVIPOSITOR LENGTH OF A CRICKET, Ecology, 76(5), 1995, pp. 1473-1482
Females of many insects possess an ovipositor, which in crickets is us
ually used to place eggs deep into the soil. Ovipositor length in fema
le striped ground crickets (Allonemobius socius) increased at higher l
atitudes and altitudes in North America, despite an opposing trend in
overall body size. Crickets inhabiting regions with short growing seas
ons and long, cold winters invariably had longer ovipositors than cric
kets inhabiting warmer regions. The only exception to this observation
was for first generation crickets in bivoltine populations where ther
e was no pattern of geographic variation in ovipositor length. However
, second generation females from these same populations always had lon
ger ovipositors than their first generation ancestors, and followed a
pattern of geographic variation similar to that found for univoltine c
rickets. Overall, these results support Masaki's (1986) model, which p
redicts that longer ovipositors will evolve in northern cricket popula
tions as an evolutionary response to the selective advantages of overw
intering eggs being oviposited deep in the soil in cold habitats. In o
rder to determine the mechanisms underlying patterns of phenotypic var
iation in the wild, and to test the hypothesis that interpopulation va
riation reflects genetic differentiation while intergeneration variati
on within a population reflects environmentally induced phenotypic pla
sticity, crickets from eight populations were brought to the laborator
y and reared through three generations in a common garden. It was foun
d that laboratory-reared populations varied significantly in oviposito
r length, with the pattern of variation mimicking that found for field
-collected crickets. This finding supports the hypothesis that geograp
hic variation in the wild has a significant genetic basis. The influen
ce of environment on ovipositor length was examined in 20 full-sib fam
ilies from a bivoltine population. Crickets reared in ''autumn-like''
conditions (30 degrees C, 15:9 h, L:D), such as might be experienced b
y second generation bivoltine crickets in the wild, were generally lar
ger and had significantly longer ovipositors than siblings reared unde
r ''spring-like'' conditions (27 degrees C, 11:13 h, L:D). This patter
n corresponds to that observed for bivoltine populations in the wild a
nd suggests the importance of environmental cues for phenotypic plasti
city, or polyphenism in ovipositor length in bivoltine populations of
this cricket.